Residue Depletion of Nitrofuran Drugs and Their Tissue-Bound

Aug 13, 2008 - An Applied Biosystems (Foster City, CA) Sciex API 2000 triple quadrupole mass spectrometer with an atmospheric pressure chemical...
0 downloads 0 Views 162KB Size
8030

J. Agric. Food Chem. 2008, 56, 8030–8034

Residue Depletion of Nitrofuran Drugs and Their Tissue-Bound Metabolites in Channel Catfish (Ictalurus punctatus) after Oral Dosing PAK-SIN CHU,*,† MAYDA I. LOPEZ,† ANN ABRAHAM,‡ KATHLEEN R. EL SAID,‡ ‡ AND STEVEN M. PLAKAS Center for Veterinary Medicine, U.S. Food and Drug Administration, 8401 Muirkirk Road, Laurel, Maryland 20708, and Gulf Coast Seafood Laboratory, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, P.O. Box 158, 1 Iberville Drive, Dauphin Island, Alabama 36528

The depletion of the nitrofuran drugs furazolidone, nitrofurazone, furaltadone, and nitrofurantoin and their tissue-bound metabolites [3-amino-2-oxazolidinone (AOZ), semicarbazide (SC), 3-amino-5morpholinomethyl-2-oxazolidinone (AMOZ), and 1-aminohydantoin (AH), respectively] were examined in the muscle of channel catfish following oral dosing (1 mg/kg body weight). Parent drugs were measurable in muscle within 2 h. Peak levels were found at 4 h for furazolidone (30.4 ng/g) and at 12 h for nitrofurazone, furaltadone, and nitrofurantoin (104, 35.2, and 9.8 ng/g respectively). Parent drugs were rapidly eliminated from muscle, and tissue concentrations fell below the limit of detection (1 ng/g) at 96 h. Peak levels of tissue-bound AMOZ and AOZ (46.8 and 33.7 ng/g respectively) were measured at 12 h, and of SC and AH (31.1 and 9.1 ng/g, respectively) at 24 h. Tissue-bound metabolites were measurable for up to 56 days postdose. These results support the use of tissuebound metabolites as target analytes for monitoring nitrofuran drugs in channel catfish. KEYWORDS: Nitrofurans; channel catfish; LC/MS-MS; depletion; bound residues

INTRODUCTION

Nitrofuran drugs (NFs, Figure 1) have been used since the 1940s in human and veterinary medicine, principally for treatment of bacterial and protozoan infections (1). Due to mutagenicity and carcinogenicity concerns, many countries, including the United States (2), have banned the use of NFs in food-producing animals. However NFs continue to be a regulatory concern, as violative residues are still being found in animal food products. Previous studies have demonstrated that parent NFs deplete rapidly in animals and that NFs are extensively metabolized to tissue-bound metabolites (3–5). As such, newer analytical methods have been directed to the determination of the intact side-chains of tissue-bound residues, instead of the parent NFs. To date, methods have been described for animal tissues (6–9), honey (10, 11), shrimp (12), eggs (13), and milk (14). Despite the availability of analytical methods, few studies have addressed the residue depletion of NFs and of their tissue-bound metabolites in aquatic animal species. Although Stehly et al. (15) examined the pharmacokinetics, tissue distribution, and metabolism of 14C-nitrofurantoin in channel catfish after intravenous and oral dosing, and Plakas et al. (16) of 14C-furazolidone * To whom correspondence should be addressed. Phone: 301-2104583. Fax: 301-210-4653. E-mail: [email protected]. † Center for Veterinary Medicine. ‡ Gulf Coast Seafood Laboratory.

10.1021/jf801398p

in channel catfish, these studies were conducted with the use of radiotracer techniques over relatively short withdrawal times. Studies investigating the persistence of bound residues over prolonged withdrawal periods and the usefulness of using bound residues as marker residues of NFs for regulatory purposes are needed. The present study examines the depletion of furazolidone, nitrofurazone, furaltadone, and nitrofurantoin, and of their tissuebound metabolites [3-amino-2-oxazolidinone (AOZ), semicarbazide (SC), 3-amino-5-morpholinomethyl-2-oxazolidinone (AMOZ), and 1-aminohydantoin (AH), respectively], from muscle of channel catfish following oral dosing. Tissue-bound metabolites were monitored for up to eight weeks after dosing by using LC-MS/MS. Residues resulting from waterborne exposures to NFs were also examined. MATERIALS AND METHODS Reagents. LC grade water was purified in-house with a Milli-Q Plus water system for use in preparation of all solutions. ACS-grade ammonium acetate (NH4OAc), anhydrous potassium phosphate dibasic (K2HPO4), N,N-dimethylformamide (DMF), polyethylene glycol (PEG, mol. wt. 200), and sodium hydroxide (NaOH) were obtained from Sigma-Aldrich (St. Louis, MO), and glacial acetic acid (HOAc) from J. T. Baker (Phillipsburg, PA). Methanol (MeOH), hexane, acetonitrile (CH3CN), and ethyl acetate (EtOAc) were high purity solvents obtained from Burdick & Jackson (Muskegon, MI). Dimethyl sulfoxide (DMSO) and reagent alcohol (containing 95% 200 proof ethanol and 5% isopropyl alcohol) were obtained from EM Science (Gibbstown, NJ).

This article not subject to U.S. Copyright. Published 2008 by the American Chemical Society Published on Web 08/13/2008

Depletion of Nitrofuran Bound Residues in Channel Catfish

J. Agric. Food Chem., Vol. 56, No. 17, 2008

8031

Figure 1. Structures of parent nitrofurans (NFs), their side-chains, and nitrophenyl derivatives. Nitrofurazone, furazolidone, furaltadone, nitrofurantoin, 1-aminohydantoin hydrochloride, semicarbazide hydrochloride, and 2-nitrobenzaldehyde (2-NBA) were purchased from Sigma-Aldrich. AOZ and AMOZ were obtained from WITEGA (Berlin, Germany). The internal standards semicarbazide-13C,15N2 (SC+3) hydrochloride, and 3-amino2-oxazolidinone-d4 (AOZ-d4) were obtained from Sigma-Aldrich while 3-amino-5-morpholinomethyl-2-oxazolidinone-d5 (AMOZ-d5) was obtained from WITEGA. In this paper, the letters NP are added to the abbreviated notation for the side chains when referring to their nitrophenyl derivatives (e.g., SC vs NPSC). Standard Solutions for Tissue-Bound Metabolites. Standard solutions (100 µg/mL) of AOZ, SC, AH, and AMOZ were prepared by dissolving appropriate weights of each solid standard in methanol. All standards are expressed as the un-ionized moiety equivalent. Each standard was weighed to the nearest 0.1 mg into respective 100 mL volumetric flask and brought to the mark with MeOH. These solutions were immediately used to prepare an intermediate mixed standard solution consisting of 2 µg/mL of each analyte by dilution with methanol. Working standard solutions (0.2 µg/mL and 0.04 µg/mL) were prepared by diluting the intermediate mixed standard solution with methanol. Internal standard solutions (AOZ-d4, AMOZ-d5, and SC+3) were prepared in a similar manner, except that smaller quantities were weighed. All standard solutions were stored at -20 °C until use. Standard Solutions for Parent NFs. Solid standards (5.0 mg) of nitrofurazone, furazolidone, furaltadone, and nitrofurantoin were weighed, dissolved in 10 mL DMF, and diluted to 50 mL with CH3CN to give 100 µg/mL stock solutions. Mixed-standard intermediate solutions (0.125-4 µg/mL) were prepared by diluting in CH3CN, and working standards (5-160 ng/mL) by dilution with water. All standard solutions were protected from light, and stored at -20 °C until use. Liquid Chromatograph (for Parent NFs). The LC system consisted of Shimadzu Corp. (Kyoto, Japan) SCL-10A system controller, LC-10AS pumps, SCL-10A autoinjector, CTO-10A column oven, and SPD-10AV UV-vis spectrophotometric detector,

with EZChrom chromatography data system. The LC column was an Inertsil ODS-3 5 µm, 4.6 × 150 mm, with a guard column (MetaGuard) of the same packing (Varian Inc., Walnut Creek, CA). Liquid Chromatograph-Tandem Mass Spectrometer (for Bound Residues). The liquid chromatography system consisted of two PerkinElmer (Norwalk, CT) Series 200 micro pumps and a Perkin-Elmer Series 200 autosampler equipped with a 100 µL loop. The LC column was an Inertsil ODS-3 5 µm, 150 × 2.1 mm with a guard column of the same packing (Varian Inc., Walnut Creek, CA). A precolumn filter (Upchurch, Oak Harbor, WA) was installed between the autosampler and the guard column. An Applied Biosystems (Foster City, CA) Sciex API 2000 triple quadrupole mass spectrometer with an atmospheric pressure chemical ionization (APCI) source in the positive ion mode was used. The protonated molecules, [M+H]+ at m/z 249 (NPAH), 209 (NPSC), 236 (NPAOZ), and 335 (NPAMOZ) were selected as the precursor ions for collision induced dissociation (CID). The following product ions were identified for selected reaction monitoring (SRM) LC/MS-MS analysis: NPAH: m/z 249 f 134, 104, 178; NPSC: m/z 209 f 166, 192, 134; NPAOZ: m/z 236 f 134, 104, 149; NPAMOZ: m/z 335 f 291, 262, 128. For the internal standards, the following transitions were selected for SRM: NPSC+3, m/z 212 f 168; NPAOZ-d4, m/z 240 f 134; NPAMOZ-d5, m/z 340 f 296. The dwell time for each monitored transition was 150 ms. The source temperature and ion spray voltage were set at 350 °C and 5.5 kV, respectively. Quantitation was performed using internal standard ratio method. Peak areas for quantitation were computed by summing the areas of the product ions of the respective analytes upon integration using PE-Sciex Analyst version 1.3.2 software. The corresponding isotopically labeled compounds were used as internal standards for AOZ, SC, and AMOZ. SC+3 was used as the internal standard for AH, as an isotopically labeled analogue was not available. A weighting of 1/x for area responses of standards provided the best linear fit of data. A negative control and a fortified control at 2 ng/g were processed together with each sample set.

8032

J. Agric. Food Chem., Vol. 56, No. 17, 2008

Dosing and Sampling. Oral Dosing. Drug-free channel catfish (Ictalurus punctatus), mean body weight 0.56 kg, were obtained from a local fish farm (Brewton, AL), and acclimated for a minimum of two weeks at the Gulf Coast Seafood Laboratory in 350-gal holding tanks. On the day before dosing, fish were randomly selected and placed in 75-gal Living Stream tanks (Frigid Units Inc., Toledo, OH) equipped with activated carbon filtration system, five fish per tank. Mean water temperature was 19 °C and pH 7.8. Dosing solutions (light protected) for each drug were prepared by solubilizing 100 mg in 10 mL of DMF, and diluting to 50 mL with PEG, to give 2 mg/mL active drug. Fish were dosed orally at 1 mg/kg body weight by delivering a gelatin capsule (size No. 00, Eli Lilly and Co., Indianapolis, IN) containing the dose solution (0.5 mL/kg body weight) and 0.25 g ground catfish feed (Purina Mills Inc., St. Louis, MO) directly into the stomachs of anesthetized (tricaine methanesulfonate, MS 222, Sigma-Aldrich) animals. Fish (n ) 5) were sampled at each of the following time points per drug: 2, 4, 8, and 12 h; and 1, 4, 7, 10, 14, 28, and 56 days after dosing. Each Living Stream tank was dedicated to a single withdrawal period per given drug and reused as necessary to accommodate all sampling time points. Fish were killed by cervical dislocation while under anesthesia. Muscle tissues (skinless fillets) were collected, homogenized, and frozen at -80 °C in 50-mL polypropylene centrifuge tubes. Bath Treatment. To examine tissue-bound residues after waterborne exposure, groups of five channel catfish were exposed in a bath treatment containing 10 mg/L of the individual NFs for 8 h. After treatment at zero-hour withdrawal time, fish were rinsed briefly and killed under anesthesia. The muscle tissue was harvested, homogenized, and stored for analysis as above. Extraction and Analysis of Parent NFs Using LC-UV. Extraction of parent NFs from muscle tissues followed the method of Rupp et al. (17). Briefly, parent NFs were extracted from the tissue with CH3CN, and the tissue lipids removed from the extracts with hexane. Extracts were evaporated by rotary evaporation, and the dried residues were redissolved in 90:10 water/CH3CN. The mixture was sonicated, centrifuged, and filtered. Extracts were analyzed for parent NFs using the following isocratic LC conditions: mobile phase, 10 mM NH4OAc in 0.02% HOAc/CH3CN (90/10); flow rate, 1 mL/min; column temperature, 40 °C; UV detector, 368 nm; and injection volume, 50 µL. The limit of detection and the limit of quantitation of parent NFs in muscle were estimated to be 1 and 2.5 ng/g, respectively. Extraction and Analysis of Tissue-Bound Metabolites Using LCMS/MS. The extraction procedure was a modification of that described for shrimp (12). In brief, channel catfish tissues (2.0 ( 0.2 g) were weighed into 50 mL polypropylene centrifuge tubes. Five mL of 70% aqueous MeOH were added, and the tubes vortex-mixed for 15 s or until the pellets were dispersed. After centrifugation at 15 °C for 5 min at 4000 rpm (∼3300g), the supernatants were decanted and discarded. The remaining pellets were sequentially washed with 5 mL each of ethyl acetate and reagent alcohol (200 proof ethanol) by vortexmixing and centrifugation. The supernatants were decanted and discarded. Fortification of tissue samples and preparation of calibration standards were performed at this step. A calibration standard curve bracketing the unknown concentrations was prepared by adding the appropriate volume of the fortification solution to 50 mL centrifuge tubes. One hundred µL of the mixed internal standard (equivalent to 2 ng/g of AMOZ-d5, 5.9 ng/g of AOZ-d4 and 8 ng/g SC+3 in tissue) was added to all samples, including the control and the standards. To the tissue samples and calibration standards were added 10 mL of 0.125 M HCl and 400 µL of freshly prepared 2-NBA solution (50 mM in DMSO). The samples were vortex-mixed for 15 s and placed in a 37 °C water bath overnight (∼14-16 h) with gentle shaking. One milliliter of 0.1 M K2HPO4 was added to the samples, and the pH of each adjusted to 7.1-7.5 by adding 0.8 M aqueous NaOH. The tissue samples (but not the calibration standards) were centrifuged at 4000 rpm for 5 min at 4 °C. The supernatants were decanted into respective 20 mL reservoirs equipped with filtering frits and attached to an SPE manifold. A slight vacuum was needed to start the flow. The filtrates were collected into 50 mL polypropylene tubes. To avoid analyte loss, the remaining pellets were washed twice with 3 mL of water. The tubes were vigorously vortex-mixed to break up the pellets and centrifuged

Chu et al.

Figure 2. Depletion of parent NFs in muscle of channel catfish after oral

dosing (1 mg/kg body weight) with furazolidone, nitrofurazone, furaltadone, and nitrofurantoin. at 4000 rpm for 5 min at 4 °C. The supernatants were decanted into their respective reservoirs and filtered. All samples (including the calibration standards) were adjusted to 20-22 mL by adding Milli-Q water. The actual volume could vary from set to set; however, all samples within the set were made to the same volume. Approximately 1.5 g of NaCl, 3 mL of reagent alcohol, and 10 mL of hexane were added. The samples were gently hand-shaken and centrifuged at 4 °C for 10 min at 4000 rpm for phase separation. The top hexane layers were removed and discarded. The remaining aqueous layers were partitioned three times with 4 mL of EtOAc by hand-shaking. After centrifugation for 5 min at 4 °C and 4000 rpm, the EtOAc layers were transferred into 15 mL polypropylene centrifuge tubes. The combined EtOAc layers were washed twice with 1 mL of Milli-Q water. The bottom aqueous layers were discarded. The remaining EtOAc layers were evaporated to dryness using a Zymark evaporator at 40 °C. To prevent analyte loss on the tube wall, EtOAc or MeOH (2 mL) was added to the samples followed by vortex-mixing and evaporation to dryness. The residues were dissolved in 200 µL of the LC mobile phase. After vortex-mixing, sonication for 2 min, and vortex-mixing again, the final extracts were filtered through 0.2 µm Gelman nylon filters into autosampler vials with 300 µL inserts. Extracts were analyzed for bound residues of NFs using the following isocratic LC conditions: mobile phase, 55% MeOH/45% 20 mM NH4OAc and flow rate, 200 µL/min. Typical injection sequence was as follows: a water blank to equilibrate the LC system, 40 µL of each calibration standard, a sample set, and the standards again. At the end of each day, the analytical and guard columns were flushed with waterMeOH (10 + 90, v/v). The limits of detection were calculated as defined in the U.S. Pharmacopeia-National Formulary (18) and estimated to be 0.1, 0.1, 0.2, and 0.1 ng/g for AOZ, SC, AH, and AMOZ, respectively. The limits of quantitation were estimated to be 0.1, 0.2, 0.3, and 0.1 ng/g for AOZ, SC, AH, and AMOZ, respectively. RESULTS AND DISCUSSION

Parent NFs were found in the muscle of channel catfish within 2 h after oral dosing (Figure 2). Peak concentrations were found at 4 h for furazolidone (30.4 ng/g), and at 12 h for nitrofurazone, furaltadone, and nitrofurantoin (104, 35.2, and 9.8 ng/g, respectively). Thereafter, parent NFs were rapidly eliminated from muscle, with concentrations e5 ng/g at 24 h. At 96 h, parent NFs were not detectable (